CS 1114: Introduction to Computing Using MATLAB and Robotics Prof. Graeme Bailey (notes modified from Noah Snavely, Spring 2009)

2 Overview  What is CS 1114? –An honors-level intro to CS using camera-controlled robots (Rovio, Sony Aibo, iRobot Create) –An alternative to CS1112 or CS1132, to fulfill your Matlab computing requirement –Formerly known as CS100R

Goals of CS1114  Give you an intuition about computation problem solving  Teach you useful (and interesting) computer science  Give you fluency in the Matlab programming environment  Have fun with robots 3

Requirements  Exposure to programming (in any language)  Some interest in math –Computer science is about much more than programming, and so is this course 4

Java or Matlab?  Both CS1110 and CS111[2,4] teach fundamental problem solving skills and computer science techniques  The destination is the same…  … but the vehicle is different 5 (inspired by Charlie Van Loan)

Robots:

7 Robots: cute but dumb  What do they know about the world around them? –Without your help, very little –Can’t even notice a bright red lightstick  Your mission: make them smarter  Lots of interesting math and computer science, some computer programming –Lots of experience with programming, even with robots, won’t give you a leg up in 1114

8 CS1114 Logistics  Lectures: Tue Thu 11:15–12:05, PHL 307  Sections: –Wed 1:25 - 2:15, Upson 317 –Wed 2:30 - 3:20, Upson 317 –Wed 3:35 - 4:25, Upson 317 –Please go to same section for the entire course  Occasional evening lectures (optional) –First lecture will be in next week (TBA)

CS1114 Logistics  CS1114 lab: Upson 317  You will soon have access to the lab and passwords for the computers  Office hours will generally be held in the lab (hours to be announced soon) 9

10 Assignments  Several mini-quizes –In class, usually at start of lecture Corollary: be on time, or write fast…  5-6 robot programming assignments with multiple parts –You will demo each part to the lab TA’s  2-3 prelims  Free-form final project (required)

What can we do with computer vision? 11

Human vision 12 Question: How many people are in this image? Source: “80 million tiny images” by Torralba, et al.

Interpreting images 13 Q: Can a computer (or robot) understand this image? A: Yes and no (mostly no)

Human vision has its shortcomings… 14 Sinha and Poggio, Nature, 1996 Credit: Steve Seitz

Human vision has its shortcomings… 15 by Ted Adelson, slide credit Steve Seitz

16

Interpreting images 17 Q: Can a computer (or robot) find the lightstick? A: With your help, yes! * (300, 100)

What is an image? 18 “lightstick1.jpg”

19 Major CS1114 Projects  From a camera, figure out the position of a bright red lightstick –Use this to guide a robot around What we see What the robot sees

What is an image?  A grid of numbers (intensity values)  Intensity values range between 0 (black) and 255 (white) 20 x y … … snoop 3D view

What is an image?  A grid of numbers (intensity values)  In Matlab, a matrix 21 [ … ; … ; … ; … ] { { x 100 matrix

Matrices in Matlab  1D matrix is often called a vector –Similar to arrays in other languages 22 A = [ ] Row vector (or 1 x 5 matrix) B = [ 10 ; 30 ; 40 ; 106 ; 123 ] Column vector (or 5 x 1 matrix) A(1) == 10 A(4) == 106

Matrices in Matlab 23 C = [ ; ; ] 3 x 5 matrix C(1,1) == ? C(2,4) == ? can also assign to a matrix entries C(1,1) = C(1,1) + 1

Image processing  We often want to modify an image by “doing something” to each pixel: 24 BlurBrighten

Brightening an image 25 Q: What does this mean in terms of matrix entries? [ ; ; ] A: Increase each element by some amount (say, 20)

Brightening an image (Take 1) 26 D = [ ] D(1) = D(1) + 20; D(2) = D(2) + 20; D(3) = D(3) + 20; D(4) = D(4) + 20; D(5) = D(5) + 20; D(6) = D(6) + 20; D(7) = D(7) + 20; D(8) = D(8) + 20; D(9) = D(8) + 20; D(10) = D(10) + 20; D(11) = D(11) + 20; D(12) = D(12) + 20; D(13) = D(13) + 20; D(14) = D(14) + 20; D(15) = D(15) + 20; x

27 Avoiding duplicate code  Programming languages are designed to make this easy –It’s a huge theme in language design –Many new programming techniques are justified by this Object-oriented programming, higher-order procedures, functional programming, etc.

Why is it a bad idea to duplicate code? 28  Hard to write  Hard to modify  Hard to get right  Hard to generalize  Programmer’s “intent” is obscured D(1) = D(1) + 20; D(2) = D(2) + 20; D(3) = D(3) + 20; D(4) = D(4) + 20; D(5) = D(5) + 20; D(6) = D(6) + 20; D(7) = D(7) + 20; D(8) = D(8) + 20; D(9) = D(8) + 20; D(10) = D(10) + 20; D(11) = D(11) + 20; D(12) = D(12) + 20;

29 Brightening an image (Take 2)  Using iteration for i = 1:12 D(i) = D(i) + 20; end D(1) = D(1) + 20; D(2) = D(2) + 20; D(3) = D(3) + 20; D(4) = D(4) + 20; D(5) = D(5) + 20; D(6) = D(6) + 20; D(7) = D(7) + 20; D(8) = D(8) + 20; D(9) = D(9) + 20; D(10) = D(10) + 20; D(11) = D(11) + 20; D(12) = D(12) + 20;  Much easier to understand and modify the code  Better expresses programmer’s “intent” D = D + 20; “Vectorized” code Usually much faster than loops

Many advantages to iteration  Can do things with iteration that you can’t do by just writing lots of statements  Example: increment every vector cell –Without knowing the length of the vector! len = length(D); % New Matlab function for i = 1:len D(i) = D(i) + 20; end 30

31 Introducing iteration into code  Programming often involves “clichés” –Patterns of code rewriting –I will loosely call these “design patterns”  Iteration is our first example

Brightening 2D images 32 C = [ ; ; ] 3 x 5 matrix for row = 1:3 for col = 1:5 C(row,col) = C(row,col) + 20; end Called a “nested” for loop

Brightening 2D images 33 for row = 1:3 for col = 1:5 C(row,col) = C(row,col) + 20 end  What if it’s not a 3x5 matrix? [nrows,ncols] = size(C) % New Matlab function for row = 1:nrows for col = 1:ncols C(row,col) = C(row,col) + 20 end

What about red pixels?  A grayscale image is a 2D array –Brightest = 255, darkest = 0 34

What about red pixels?  A color image is 3 different 2D arrays –For red/green/blue values (RGB) –We provide a way to create these 3 arrays  Why are there 3? 35 =

What about red pixels?  Example colors: red(1,1) == 255, green(1,1) == blue(1,1) == 0 red(2,1) == 100 == green(2,1) == blue(2,1) red(3,1) == 0 == green(3,1) == blue(3,1) red(3,1) == 255 == green(3,1), blue(3,1) == 0 36

For next time  ……….. Wait* and see** ……… !!!*** ______________ * Assuming you know how to ‘wait’ ** assuming you know how to ‘see’ *** assuming you know why waiting helps seeing 37